Explorer

Header
Our Questions

Inspiration

For most people, planning a trip to travel to a new place or new country can be exhausting. So much information needs to be collected, and it can take hours to prepared the perfect trip that meets the preferences and budget of the traveler. We want to make finding and organizing that information as easy as possible, making it easier for people all over the world come together, have a good travel experience, and feel comfortable in new places.

What it does

Our website asks the user for some basic information about their background and goals for their journey. These questions include information about location, budget, language preferences, nationality, religious preferences, who they are traveling with, what they like to do, etc. From this information, a response is generated using Google Gemini that provides trip information/advise based on the user's preferences. The site auto-detects what language the user uses to input their information, and outputs information in the same language. Our goal is to quickly provide information that is concise, budget friendly, and matches with the users' preferences. This way we can make travel more accessible to all, and ultimately connect the world.

How we built it

We designed our app with the user's input at its core. Initially, we gather specifics from the user about their trip preferences and budget. This information is then crafted into a prompt for our AI, Gemini, to generate a custom itinerary. The final step involves organizing Gemini's output into a concise list of suggested activities, places to visit, and accommodation options for the user's trip.

Challenges we ran into

The majority of our group had never participated in a hackathon, and we all had little experience with Software Development and using API keys. Our team was also using GitHub in a group setting, so in the beginning there was a bit of a learning curve. We had trouble initially getting the repository set up and working on everyone's computers.

Accomplishments that we're proud of

1) We managed to get through the process with only one merge conflict that caused significant issues.

2) We were also proud of what we were able to accomplish in the short time span of the hackathon. Considering most of us had little to no experience creating websites, completing out-of-class projects, or using API keys to integrate AI into our projects.

What we learned

During this project, we strengthened our knowledge of Frontend and Backend development. We also learned how to effectively communicate and divide up tasks in a group setting.

What's next for Explorer

Given more time, we definitely would have formatted the AI output to be easier to read. Right now, the AI output the correct information but is not displayed as well as it could be.

We also could have improved the AI prompting to find even better results. One way would be to instruct the AI to look through reviews for each of the itinerary items and rank them based on the User’s interests. We could also have instructed the AI to initially produce a larger list of itinerary items and then choose the best ones from that list.

Other possible features could include flight/travel plans, possible risks associated with the trip, and a separate service to help the user choose where they want to travel based on their interests.